Involvement of CYP2D6 in the in vitro metabolism of amphetamine, two N-alkylamphetamines and their 4-methoxylated derivatives

Article abstract of DOI:10.1080/004982599238344

1. Amphetamine (AM) and five amphetamine derivatives, N-ethylamphetamine (NEA), N-butylamphetamine (NBA), 4-methoxyamphetamine (M-AM), 4-methoxy-N-ethylamphetamine (M-NEA) and 4-methoxy-N-butylamphetamine (M-NBA) were incubated with microsomal preparations from cells expressing human CYP2D6 to determine whether the enzyme was capable of catalyzing the direct ring oxidation of all substrates; the N-dealkylation of NEA, NBA, M-NEA and M-NBA; and the O-demethylation of M-AM, M-NEA and M-NBA. 2. None of the six compounds examined was N-dealkylated to any extent. 3. The only metabolites produced from AM, NEA and NBA were the corresponding ring 4-hydroxylated compounds, and the rates of formation were low. 4. All ring 4-methoxylated substrates were efficiently O-demethylated by CYP2D6 to their corresponding phenols. The size of the N-alkyl group influenced the rates of formation of these phenolamines. In contrast to reported findings with 2- and 3-methoxyamphetamines, none of the 4-methoxyamphetamines was ring-oxidized in the CYP2D6 enzyme system to 2- or 3-hydroxy-4-methoxyamphetamines or to dihydroxy-amphetamines.

Full text of DOI:10.1080/004982599238344

xenobiotica , 1999, vol . 29, no. 7, 719±732
Involvem ent of CYP2D6 in the
metabolism of
in vitro
amphetam ine, two N-alkylamphetam ines and their
4-methoxylated derivatives
MIMI V. BACH‹, R. T. COUTTS‹* and G. B. BAKERŒ
‹
Neurochemical Research Unit, Faculty of Pharmacy and Pharmaceutical Sciences, and
Department of Psychiatry, University of Alberta, Edmonton, Alberta T6G 2N8,
Œ
Canada
Received 27 June 1998
1. Amphetamine (AM) and ®ve amphetamine derivatives, N-ethylamphetamine
(NEA), N-butylamphetamine (NBA), 4-methoxyamphetamine (M-AM), 4-methoxy-N-
ethylamphetamine (M-NEA) and 4-methoxy-N-butylamphetamine (M-NBA) were incu-
bated with microsomal preparations from cells expressing human CYP2D6 to determine
whether the enzyme was capable of catalyzing the direct ring oxidation of all substrates; the
N-dealkylation of NEA, NBA, M-NEA and M-NBA ; and the O-demethylation of M-AM,
M-NEA and M-NBA.
2. None of the six compounds examined was N-dealkylated to any extent.
3. The only metabolites produced from AM, NEA and NBA were the corresponding
ring 4-hydroxylated compounds, and the rates of formation were low.
4. All ring 4-methoxylated substrates were eæciently O-demethylated by CYP2D6 to
their corresponding phenols. The size of the N-alkyl group in¯uenced the rates of
formation of these phenolamines. In contrast to reported ®ndings with 2- and 3-
methoxyamphetamines, none of the 4-methoxyamphetamines was ring-oxidized in the
CYP2D6 enzyme system to 2- or 3-hydroxy-4-methoxyamphetamines or to dihydroxy-
amphetamines.
Introduction
Our interest in the in vitro metabolism of centrally active drugs with expressed
human cytochrome P-450 (CYP) enzymes initially involved the tricyclic anti-
depressants, imipramine (IMI) (Su et al. 1993) and amitriptyline (AMI) (Coutts et
al. 1997). It is our wish to report a related study on the metabolism of amphetamine
(AM), two N-alkylated amphetamines and their 4-methoxylated derivatives in an
expressed human CYP2D6 enzyme system. Like IMI and AMI, these drug
substrates possess aromatic rings attached by alkyl chains to an aliphatic amine
group and could be expected to undergo metabolic ring oxidation to phenolic
products, and possibly N-dealkylation. In addition, the ring methoxylated com-
pounds are good candidates for O-demethylation.
Except for N-methylamphetamine (NMA) and some of its ring-substituted
derivatives, few studies on the metabolism of N-alkylated amphetamines in man
have been reported. From those that have, it is apparent that both N-dealkylation
and ring hydroxylation of N-alkylamphetamines occur only to a small extent in man
(Smith and Dring 1970, Gorrod 1973). In vivo ring hydroxylation of AM to 4-
hydroxyamphetamine (HO-AM), for example, is a minor metabolic pathway in man
C
(
5 % in a 24-h urine), whereas ring oxidative metabolism of NMA is more
!
* Author for correspondence ; e-mail : ronald.coutts ualberta.ca
Xenobiotica ISSN 0049±8254 print}ISSN 1366±5928 online ’ 1999 Taylor & Francis Ltd
http:}}www.tandf.co.uk}JNLS}xen.htm
http:}}www.taylorandfrancis.com}JNLS}xen.htm

720
M. V. Bach et al.
extensive and yields 4-hydroxy-N-methylamphetamine (HO-NMA) and HO-AM
in yields of 15 and 2±3 % respectively. Ring hydroxylation of pondinil [(N-3-
chloropropyl)amphetamine; PD) is even more pronounced ; the quantities of 4-
C
6 % respectively (Williams et al. 1973, Caldwell 1976). From these data, it
hydroxy-PD and HO-AM excreted in a 24-h urine in subjects taking PD were
C
30
and
would seem that an increase in lipophilicity as a result of N-substitution results in
an increase in the extent of metabolic ring hydroxylation.
Previous studies on the eåects of the alkyl chain length in simple N-
alkylamphetamines on in vivo metabolic ring hydroxylation in rats and man (Coutts
et al. 1976, 1978, Coutts and Dawson 1977) demonstrated that AM and NMA
formed only 4-hydroxylated metabolites whereas N-ethylamphetamine (NEA), N-
(n-propyl)amphetamine (NPA) and N-(n-butyl)amphetamine (NBA) in rat, and
NPA in man, formed 4-hydroxylated and 3,4-dihydroxylated metabolites. The
dihydroxylated products were further metabolized by catechol-O-methyltransferase
(COMT)-catalysed O-methylation to the corresponding 3-methoxy-4-hydroxy-N-
alkylated amphetamines (Coutts and Jones 1982). In an in vivo study in rat (Coutts
et al. 1978), the ratio of urinary 3-methoxy-4-hydroxy : 4-hydroxy metabolites
isolated increased as the length of the N-alkyl substituent increased in size from
=
= =
0.139) to n-propyl (ratio 0.179) to n-butyl (ratio 0.435).
ethyl (ratio
Various studies on the metabolism of methoxyamphetamines in vivo (Kitchen et
al. 1979, Midha et al. 1981, Cho et al. 1990, Lin et al. 1992, Hiratsuka et al. 1995,
Chu et al. 1996) and in vitro (Beckett and Midha 1974, Coutts et al. 1994, Kumagai
et al. 1994, Geertsens et al. 1995) have shown that these compounds can undergo O-
demethylation, ring hydroxylation and N-dealkylation to various extents. The site
of ring hydroxylation, if it does occur, depends on the ring location of the methoxy
group. Whether the nature of the N-alkyl substitutent in¯uences ring hydroxylation
is not known.
Many methoxyamphetamines possess hallucinogenic properties, and the most
potent compounds are 4-methoxylated (Tseng et al. 1976). 4-Methoxyamphetamine
(M-AM), for example, is ®ve times as potent as mescaline (Shulgin et al. 1969). Since
abusers of methoxyamphetamines often ingest mixtures of drugs, there is no doubt
that drug±drug interactions resulting in the inhibition of the metabolism of 4-
methoxyamphetamines and prolongation of their stimulant eåects, are likely to
occur. Studies on compounds that will identify the cytochrome P450 (CYP)
enzymes involved in O-demethylation, ring hydroxylation and N-dealkylation are
therefore relevant.
Some in vitro studies have suggested that metabolic O-demethylation of
methoxyamphetamines is mediated by CYP2D6 in man (Coutts et al. 1994, Tucker
et al. 1994, Geertsen et al. 1995) and by related CYP2D enzymes in rat (Kumagai et
al. 1994). It has also been claimed that N-dealkylation of some analogues of NMA
is catalysed by CYP2D6 in man (Kumagai et al. 1994, Geertsen et al. 1995,
Hiratsuka et al. 1995) and by CYP2B and CYP2D enzymes in rat (Kumagai et al.
1994). CYP2D6 is also known to catalyse the ring hydroxylation of many amines
(Coutts et al. 1994, Masubuchi et al. 1994, Narimatsu et al. 1994, Nielsen et al. 1994,
Ono et al. 1995, Coutts et al. 1997).
In the present study, the in vitro metabolism of the three amphetamines, AM,
NEA and NBA, and three ring-methoxylated amphetamines, M-AM, 4-methoxy-
N-ethylamphetamine (M-NEA) and 4-methoxy-N-(n-butyl)amphetamine (M-
NBA), was investigated using a forti®ed CYP2D6 preparation with the following

CYP2D6 and in vitro metabolism
721
Figure 1. Structures of the amphetamines investigated and their 4-hydroxylated metabolites.
*M-NMA, internal standard.
objectives in mind : (1) to determine whether ring hydroxylation of AM, NEA and
NBA occurred and whether the size of the N-substituent in¯uenced the extent of
that hydroxylation, and if ring hydroxylation did occur, to determine whether any
catecholamine metabolites were formed ; (2) to determine whether any ring oxidation
of M-AM, N-NEA and M-NBA occurred before or after O-demethylation, and
whether the length of the N-substituent had any in¯uence on the CYP2D6-
catalysed ring oxidation ; and (3) to determine whether CYP2D6 catalysed N-
dealkylation of NEA, NBA, M-NEA and M-NBA to any extent. Metabolic N-
dealkylation is generally not associated with CYP2D6, but some examples of
CYP2D6-catalysed N-dealkylation have been recognized, including ami¯amine
(Alvan et al. 1984), AMI (Coutts et al. 1997), deprenyl (Grace et al. 1994),
desmethylcitalopram (Sindrup et al. 1993), 1-methyl-4-phenyl-1,2,3,6-tetrahydro-
pyridine (MPTP) (Coleman et al. 1996), mianserin (Dahl et al. 1994) and trazodone
(Yasui et al. 1995). The structures of substrates and putative metabolic products of
the amphetamines examined are provided in ®gure 1.
M aterials and m ethods
Chemicals and reagents
All chemicals purchased from commercial sources were of analytical grade. All amphetamine
derivatives were racemic mixtures.
4-Methoxy-N-(n-butyl)amphetamine (M-NBA).HCl. 4-Methoxyphenylacetone (3 g) and a slight
excess of n-butylamine were added to 50 ml dry ethanol containing 0.15 g platinum oxide. The mixture
was hydrogenated at 35 psi for 18±24 h, catalyst was removed by ®ltration and solvent was evaporated
under vacuum. The residue was dissolved in dry ether and treated with a solution of HCl in ether until
no further precipitation occurred. This product was recrystallized from ethanol as a colourless solid,
H
ClNO : C, 65.21 %, H, 9.38 %, N, 5.43 % ; found: C, 65.41 %,
m.p. 176±177ÊC. Anal. calcd. for C
"% #%
H, 9.24 %, N, 5.28 %.

722
M. V. Bach et al.
H
ClNO :
4-Methoxy-N-ethylamphetamine (M-NEA).HCl, m.p. 158±159ÊC (anal. calcd. for C
"# #!
C, 62.74 %, H, 8.77 %, N, 6.10 % ; found : C, 62.79 %, H, 8.65 %, N, 5.84 %), and 4-methoxy-N-
H
8.40 %, N, 6.49 % ; found: C, 61.28 %, H, 8.37 %, N, 6.61%), were obtained by the same procedure
ClNO : C, 61.25 %, H,
methylamphetamine (M-NMA).HCl, m.p. 178±179ÊC (anal. calcd. for C
"" ")
except that ethylamine and methylamine respectively, were substituted for n-butylamine.
4-Hydroxy-N-(n-butyl)amphetamine (HO-NBA).HCl, m.p.
4-hydroxy-N-ethyl-
158±159ÊC,
amphetamine (HO-NEA).HCl, m.p. 159±160ÊC, and N-(n-butyl)amphetamine (NBA).HCl, m.p 166±
168ÊC were prepared by previously reported methods (Coutts et al. 1976, 1978). N-ethylamphetamine
H
ClN : C, 66.15 %, H, 9.08 %, N, 7.01 % ; found: C,
(NEA).HCl, m.p. 148±150ÊC (anal. calcd. for C
"" ")
66.28 %, H, 9.00 %, N, 6.89 %) was prepared in 88 % yield by the procedure used for N-(n-
butyl)amphetamine hydrochloride except that butylamine was replaced by ethylamine. Amphetamine
(AM) sulphate, 4-methoxyamphetamine (M-AM).HCl and 4-hydroxyamphetamine (HO-AM). HBr were
gifts from the former Smith, Kline & French Laboratories (Philadelphia, PA, USA). Other chemicals
were obtained from commercial sources: NADP+ sodium salt from yeast, -glucose 6-phosphate-
d
monosodium salt (G6P) and glucose 6-phosphate dehydrogenase (G6PD) type XII from Torula yeast
(Sigma Chemical Co., St Louis, MO, USA); acetic anhydride and pyridine (Fisher Scienti®c, Ottawa,
Canada); KHCO , K CO and quinidine sulphate (BDH, Toronto, Canada). Dichloromethane (CH Cl )
$
#
$
#
#
and diethyl ether were distilled before used. The K CO solution used in this study was a 25 % w v
}
#
aqueous solution. The buåer used in metabolic reactions was aqueous potassium phosphate buåer
$
m
(100 m ; pH 7.4).
Kimax glass culture tubes (16 mm o.d.3 100 mm) purchased from Fisher Scienti®c were used in
metabolism studies.
Puri®ed human CYP2D6 isozyme
The human CYP2D6 microsomal product used in this study was derived from a human AHH-1
–
purchased from Gentest Corporation (Woburn, MA, USA). Total protein content was 10 mg ml in
TK 1
cell line transfected with complementary DNA that encoded human CYP2D6 and was
}
}
m
100 m potassium phosphate (pH 7.4), and CYP2D6 content was 170 pmol mg protein. Control
}
microsomes that were obtained from the same human cell line that had not been transfected with speci®c
cDNA, were purchased from the same source. These microsomes contained no CYP2D6 but possessed
a low level of a cytochrome P450 that was inducible with polycyclic aromatic hydrocarbons (data not
shown).
GC assay procedure
Instrumental analysis. The gas chromatograph was an HP 5730A instrument (Hewlett Packard, PA,
USA) equipped with a nitrogen±phosphorus detector and an HP-3396A integrator. It contained a DB-
l
5 fused capillary column, 13 m3 0.25 mm i.d.3 0.25 m ®lm thickness (J&W Scienti®c, Folsom, CA,
USA). The carrier gas was ultrapure helium (Union Carbide, Edmonton, Canada) and its ¯ow rate was
adjusted to maintain a column head pressure of 10 psi. Make-up gas at the detector was a mixture of
hydrogen (3 ml min) and air (80 ml min). The injector and detector temperatures were 260 and 310ÊC
}
}
respectively. Chromatograms were recorded with an HP 3396A integrator, and peak areas were
measured. An oven temperature program was used in which the starting temperature (130ÊC) was raised
at 8ÊC min to 255ÊC and held at this temperature until all peaks had emerged.
}
Derivatization procedures
Anhydrous acetylation. The dried residue obtained at the end of each metabolic reaction (described above)
l
l
was mixed with acetic anhydride (50 l) and pyridine (25 l) in a test tube which was capped tightly with
screw cap before insertion in a metal heating module at 50ÊC for 30 min. The derivatized sample was
allowed to cool to room temperature before the tube was opened and its contents evaporated to dryness
l
at room temperature under a stream of dry nitrogen. The residue was dissolved in toluene (100 l) and
l
aliquots (2 l) were analysed by GC as described above.
Aqueous acetylation. The metabolism incubation reaction (described above) was terminated by cooling in
an ice bath and basifying with solid KHCO (350 mg). Internal standard (M-NMA ; 18.24 nmol) was
$
added followed by acetic anhydride (300 l), and acetylation was allowed to proceed for 25 min at room
temperature. The reaction mixture was extracted with the mixed organic solvent (CH Cl :diethylether;
l
#
#
11: 14 v v ; 3 ml3 3). The extracts were combined and evaporated under a stream of dry nitrogen and the
}
l
residue was reconstituted in toluene (100 l). Aliquots (2 l) of this solution were used for GC analysis
l
as described above.
Quantitative analysis of metabolites. Calibration curves for M-NBA and HO-NBA were prepared by

CYP2D6 and in vitro metabolism
723
adding internal standard (M-NMA.HCl; 18.24 nmol) and varying amounts of analytes (3.31±49.57 nmol
M-NBA.HCl ; 1.60±16.02 nmol HO-NBA) to separate 0.5 ml vols incubation mixture, prepared as
described below for the metabolism of M-NBA. Each mixture was heated for 5 min at 37ÊC, then cooled
l
in an ice bath before the addition of 25 % K CO (100 l) and CYP2D6 microsomal protein (0.20 mg).
#
$
The mixture was then extracted and derivatized using the same procedures as those described for the
substrate metabolism reaction. The ratios of M-NBA M-NMA peak areas and ratios of HO-NBA M-
NMA peak areas were plotted against the quantities of M-NBA and HO-NBA respectively.
}
}
Calibration curves for NBA, M-NEA, NEA, M-AM, AM, HO-NEA and HO-AM were prepared by
the same procedure as that described for the preparation of the calibration curves of M-NBA and HO-
NBA above, except that M-AM and AM were acetylated in aqueous medium.
Analysis of kinetic data
Apparent Michaelis±Menten parameters Km and Vmax were calculated using equation 1 by a least-
squares regression analysis of reaction velocities versus substrate concentration curves provided by the
computer program `WinNonLin’ (v1.0 ; Scienti®c Consulting, Inc.). The appropriateness of the ®t was
determined by visual inspection of residual patterns and residual sums of squares.
Vmax \S
=
V
.
(1)
Km 1S
In vitro metabolic experiments
NADPH-generating system. Separate stock solutions of each component were freshly prepared before
+
l
MgCl .6H O in phosphate buåer; and G6PD stock solution: 50 U ml of G6PD in phosphate buåer.
l
l
l
l
l
g l
use as follows: 20
g
l NADP in phosphate buåer; 20
g
l G6P in phosphate buåer; 6.7
}
}
}
}
#
#
Metabolic reactions of 4-methoxy-N-alkylamphetamines
Metabolism of M-NBA. A typical incubation mixture contained M-NBA (10.09 nmol), MgCl .6H O
#
(50 l stock solution), NADP (25 l solution), G6P (25 l solution), G6PD (10 l solution) and
#
+
l
phosphate buåer to a volume of 460
l
l
l
l
l. The mixture was pre-incubated at 37ÊC for 5 min in a shaking
Magni Whirl‡ constant temperature water bath (Blue M Electric Co., Blue Island, IL, USA). The
l
enzyme reaction was started by the addition of CYP2D6 (0.20 mg). Additional buåer solution (20 l) was
used to rinse the protein container and this washing was also added to the incubation reaction. The ®nal
solution (0.5 ml) was incubated at 37ÊC for 15 min. The reaction was terminated by cooling in an ice
l
bath, followed by the addition of K CO solution (100 l), which increased the pH of the incubation
#
$
mixture to between 11 and 12. M-NMA (18.24 nmol) was added as an internal standard and the resulting
solution was extracted with a mixture of dichloromethane and diethyl ether (11 : 14 ; 3 ml3 3) by
vortexing vigorously for 1 min, shaking mechanically for 5 min and centrifuging for 5 min. The organic
layer was transferred to a clean tube and evaporated to dryness under a stream of nitrogen at room
temperature. The residue was acetylated under anhydrous conditions as described above and the
acetylated sample was analysed using the GC assay procedure.
Time-course study of M-NBA metabolism. A mixture of M-NBA (40.36 nmol), CYP2D6 (0.80 mg), stock
+
l
l
l
l
solutions of MgCl \6H O (200 l), NADP (100 l), G6P (100 l), G6PD (40 l) and buåer to a volume
#
#
l
of 2 ml was incubated as described below, except that samples (250 l) were withdrawn at various time
intervals: 5, 10, 15, 20, 25 and 30 min. Each sample was cooled in an ice bath and basi®ed by an addition
l
l
of K CO solution (50 l). Buåer (300 l) and internal standard (M-NMA ; 9.12 nmol) were added before
#
$
extraction. Dried extracts from each incubation were acetylated under anhydrous conditions as described
above and acetylated samples analysed using the GC assay procedure.
Kinetic study of M-NBA. The metabolism of M-NBA was repeated as described above except that
varying quantities of M-NBA.HCl (3.10, 4.97, 10.09, 17.08, 29.49, 49.67, 55.89 and 68.31 nmol) were
used. Each extract was acetylated under anhydrous conditions (see above) and acetylated samples were
analysed using the GC assay procedure.
Eåect of varying enzyme concentration on the metabolism of M-NBA. The metabolism of M-NBA was
repeated as described above except that diåerent quantities of CYP2D6 microsomal protein (0.1, 0.2, 0.3,
0.4 and 0.5 mg) were used. The ®nal dried extracts were acetylated under anhydrous conditions (see
above), and the acetylated samples were analysed using the GC assay procedure.
M-NBA inhibition study. The metabolism of M-NBA (10.09 nmol) with CYP2D6 (0.20 mg) as described

724
M. V. Bach et al.
above was repeated, except that quinidine (5.05 nmol) was added to the medium just before the enzymatic
reaction was initiated by the addition of CYP2D6 solution.
Control metabolism of M-NBA. The metabolism of M-NBA was repeated as described above except that
the CYP2D6 enzyme solution was replaced with control microsomal protein (0.20 mg).
Metabolism and kinetic study of M-NEA. The incubation of various quantities of M-NEA (2.50±
68.4 nmol) with CYP2D6 (0.20 mg) at 37ÊC for 15 min was performed exactly as described for the
metabolism of M-NBA. The ®nal residue, which contained substrate and metabolites, was acetylated
under anhydrous conditions (see above), and the dried acetylated product was reconstituted in toluene
l
l
(100 l). Aliquots (2 l) were analysed by GC using the GC assay procedure.
Metabolism and kinetic study of M-AM. Various quantities of M-AM (2.50±68.4 nmol) were individually
metabolized with CYP2D6 (0.20 mg) exactly as described for the metabolism of M-NBA. The
incubation mixture was acetylated under aqueous conditions (see above). The dried acetylated product
l
was dissolved in toluene (100 l) and aliquots (2 l) were analysed by GC using the GC assay procedure.
l
Metabolic reactions of N-alkylamphetamines
Metabolism of NBA. NBA (15.09 nmol) was metabolized with CYP2D6 (0.5 mg) as described for the
metabolism of M-NBA, except that the metabolism mixture was incubated for 2 h at 37ÊC and the
quantity of internal standard, M-NMA, used was 9.12 nmol. The residue was acetylated under
l
anhydrous conditions as described above. The dried acetylated sample was dissolved in toluene (100 l)
l
and a 2 l aliquot was analysed by GC using the GC assay procedure.
This incubation was also performed in the presence of the CYP enzyme inhibitor, quinidine
(5.05 nmol).
Kinetic study of NBA. The metabolism of NBA was repeated using various quantities of NBA (0.9, 2.02,
3.03, 4.98, 8.02, 11.05, 14.97 and 19.60 nmol incubation). Samples were acetylated under anhydrous
conditions and analysed as described above.
}
Metabolism of NEA. This substrate was metabolized, and products were acetylated and analysed by GC
as described for the metabolism of NBA.
Metabolism of AM. This substrate was metabolized as described for the metabolism of NBA. After
incubation, internal standard (M-NMA, 9.12 nmol) was added and the resulting mixture was acetylated
l
under aqueous conditions (see above). The dried acetylated product was dissolved in toluene (100 l) and
l
aliquots (2 l) were analysed by GC using the GC assay procedure.
Results
In this study, all incubation mixtures containing amphetamines and their
metabolites were derivatized with acetic anhydride before GC analysis. Two
derivatization procedures, aqueous and anhydrous acetylation, were employed.
Aqueous acetylation was used at room temperature to derivatize AM and M-AM
incubation mixtures before their extraction. This procedure converted basic and
amphoteric compounds to neutral acetates that could be eæciently extracted into
organic solvents. When the same aqueous acetylation procedure was performed on
the N-butyl- and N-ethyl-amphetamines, complete acetylation of substrates and
metabolites was never attained. In these instances, acetylations were satisfactory
only when performed under anhydrous conditions.
When 4-methoxy-N-alkylated amphetamines were incubated with expressed
CYP2D6 in a medium containing an NADPH generating system, the corresponding
4-hydroxy-N-alkylamphetamines were formed at diåering rates that depended on
the chain length of the N-alkyl substituent. When M-NBA was the substrate, the
formation of HO-NBA was a linear process over a 30-min incubation (®gure 2), at
which time 90 % of M-NBA had already been O-demethylated to HO-NBA. Only

CYP2D6 and in vitro metabolism
725
Figure 2. Time-course study of M-NBA metabolism. Results are the mean of two values that diåered
!
by
5 %. HO-NBA (+ ); M-NBA ( ).
U
one product, HO-NBA, was detected in the GC trace of the incubation extract
(®gure 3). The formation of HO-NBA also increased linearly as the amount of
CYP2D6 used was increased from 0.1 to 0.5 mg per incubation (®gure 4). These
data indicate that M-NBA is a very good substrate for CYP2D6.
Experiments were conducted in which a constant quantity of CYP2D6 was
incubated with varying amounts of M-NBA ranging from 3.10 to 68.31 nmol
}
incubation to determine apparent Km and Vmax . It can be deduced from ®gure 5
"
that CYP2D6 was saturated when the M-NBA concentration was 60 lm. The
formation of HO-NBA exhibited simple single enzyme Michaelis±Menten
=
kinetics and resulted in 11.9‰ 1.81 lm (mean‰ SD ; n 3) and 121.5 ‰ 4.13 nmol
}
h mg protein for K and Vmax respectively. No O-demethylation of M-NBA
}
m
occurred when the control enzyme preparation was used in place of CYP2D6. O-
demethylation was also completely inhibited when the incubation medium also
contained quinidine (10 lm) (data not shown).
When M-NEA and M-AM were individually incubated with CYP2D6 under
conditions identical to those used for M-NBA, both substrates were also readily O-
demethylated to their corresponding ring hydroxylated metabolites, HO-NEA and
HO-AM respectively. The calculated intrinsic clearance values of these three
substrates decreased as the length of the N-alkyl group decreased (table 1).
The incubation of N-alkylated amphetamines, NBA, NEA and AM with an
expressed CYP2D6 enzyme preparation under the same conditions as those used to
metabolize the analogous 4-methoxy-N-alkylated amphetamines (0.2 mg CYP
enzyme preparation incubation ; incubation, 15 min) provided only trace amounts
}
of ring-hydroxylated metabolites. However, when the concentration of the CYP
enzyme preparation was increased to 0.5 mg incubation and the incubation was
}
extended to 2 h, ring 4-hydroxylated amphetamine metabolites were obtained. The

726
M. V. Bach et al.
Figure 3. GC traces of acetylated dried extracts of M-NBA incubated with : (a) control microsomal
protein; (b) CYP2D6 enzyme preparation ; and (c) CYP2D6 enzyme preparation to which
quinidine was added. M-NBA, 10.09 nmol; quinidine, 5.05 nmol ; protein solution, 0.20 mg.
Incubation time, 15 min; incubation volume, 0.5 ml.
products, 4-OH-NBA, 4-OH-NEA and 4-OH-AM respectively (®gure 1), and their
acetylated derivatives had GC retentions that were identical to those of authentic
reference compounds. The amount of HO-AM formed was less than one-third of
the amounts of HO-NBA and HO-NEA formed (table 2).
A kinetic study of NBA’s metabolism to 4-OH-NBA was performed, and
apparent Km and Vmax were calculated. The Michaelis±Menten plot of the ring

CYP2D6 and in vitro metabolism
727
Figure 4. Eåects of enzyme concentration on the yield of HO-NBA obtained by metabolism of
M-NBA. Incubation time, 15 min; volume of incubation, 0.5 ml; M-NBA, 10.09 nmol. Data
=
represent means‰ SD (n 3).
Figure 5. Kinetic saturation curve of M-NBA metabolism mediated by CYP2D6. Incubation time,
=
15 min. M-NBA, 6.2±136.6 nmol ml. Data are means‰ SD (n 3).
}
hydroxylation of NBA by CYP2D6 is provided in ®gure 6, and calculated metabolic
=
constants were 3.68‰ 0.30 lm (mean‰ SD ; n 3) and 0.62‰ 0.01 nmol h mg
} }
protein for Km and Vmax respectively. The 4-hydroxylation of NBA mediated by
CYP2D6 was completely inhibited by the inclusion of quinidine in the metabolism
reaction (data not shown). Intrinsic clearances (Clint ) of M-NBA and NBA are listed
in table 3.

728
M. V. Bach et al.
Table 1. Apparent Km, Vmax and Clin for the O-demethylation of M-NBA, M-NEA and M-AM
mediated by CYP2D6.
t
Clin
t
l
( l min mg
Apparent Km
lm
Vmax (nmol h
mg protein)
} }
}
}
Metabolic reaction
(
)
protein)
M-NBA ! HO-NBA
M-NEA ! HO-NEA
M-AM ! HO-AM
11.90‰ 1.81
19.52‰ 1.45
29.26‰ 4.41
121.51‰ 4.13
107.65‰ 4.22
107.42‰ 9.64
170.4
91.91
61.16
Values represent mean (‰ SD) of three determinations.
Table 2. Yield of aromatic hydroxylation metabolites.
Substrate
(30.18 nmol ml
incubation)
Metabolites (nmol h mg
protein)
}
} }
NBA
NEA
AM
HO-NBA : 0.58‰ 0.04
HO-NEA : 0.55‰ 0.03
HO-AM : 0.16‰ 0.02
Incubation time, 2 h. Protein, 0.50 mg incu-
}
bation.
Values represent means of three determinations
(‰ SD).
Figure 6. Kinetic saturation curve of NBA metabolism mediated by CYP2D6. Incubation time, 2 h.
lm
=
. Data are means‰ SD (n 3).
Concentration of NBA, 1.8±39.2
Table 3. Apparent Km, Vmax and Clin for M-NBA and NBA metabolism by CYP2D6.
t
Clin
t
l
( l min mg
Apparent Km
lm
Vmax (nmol h
mg protein)
} }
}
}
Metabolic reaction
(
)
protein)
NBA ! HO-NBA
M-NBA ! HO-NBA
3.68‰ 0.30
11.9‰ 1.81
0.62‰ 0.01
121.5‰ 4.13
2.81
170.4
Values represent mean (‰ SD) of three determinations.

CYP2D6 and in vitro metabolism
729
Figure 7. Structures of 2-, 3- and 4-methoxyamphetamine analogues and their metabolites in man.
Discussion
This study has shown that the 4-methoxy-N-alkylamphetamines, M-NBA, M-
NEA and M-AM, all readily undergo O-demethylation to their corresponding 4-
hydroxy-N-alkylamphetamines when incubated with a CYP2D6 enzyme system.
The rates of formation of these phenolic amines are in¯uenced by the size of alkyl
!
!
groups : hydrogen ethyl butyl (table 1). CYP2D6 did not catalyse the ring
hydroxylation of any of the three 4-methoxyamphetamines selected for this study
(®gure 3). This result contrasts with those obtained when 2-methoxyamphetamine
and methoxyphenamine (a 2-methoxyamphetamine derivative) were metabolized in
vitro with microsomal CYP2D6 or with CYP2D6-transfected intact human cells
(Coutts et al. 1994, Geertsens et al. 1995) ; these substrates were not only ring-O-
demethylated, but also were ring-hydroxylated, mainly on C-5 but, to a small
extent, on C-3. This inability to ring hydroxylate 4-methoxyamphetamines should
also be compared with reported in vivo metabolic results of studies on 3-
methoxyamphetamine. This substrate was O-demethylated (CYP enzyme not

730
M. V. Bach et al.
identi®ed) in good yield to 3-hydroxyamphetamine, but only one ring-hydroxylated
urinary metabolite, 4-hydroxy-3-methoxyamphetamine, was isolated as a minor
metabolite (Midha et al. 1981). These observations show that the extent to which
ring hydroxylation occurs depends on the ring location of the methoxy group. If it
does occur, the major sites of ring oxidation will be those para and, to a much lower
extent, ortho to the methoxy group. The para position is much less sterically
hindered (®gure 7). The current observation is in agreement with an earlier report
that no ring hydroxylated metabolites were detectable in the urine of man who had
ingested 4-methoxyamphetamine (Kitchen et al. 1979).
Of the three 4-methoxyamphetamines examined in the present study, M-NBA
was a very good substrate for CYP2D6. Km , Vmax and Clin for the three substrates
t
are compared in table 1. The size of the N-alkyl group in¯uenced the extent of O-
demethylation. As the size of the N-alkyl group increases, so does lipid solubility,
and substrate concentration at the active site of the enzyme is a function of a
substrate’s lipid solubility (Duncan et al. 1983). Although CYP2D6 catalyses the N-
dealkylation of some amine compounds, no N-dealkylation was observed in the
present study.
It was clear that CYP2D6 catalysed the ring hydroxylation of NBA, NEA and
AM as demonstrated by the formation of their 4-hydroxy N-alkylamphetamines and
by the inhibition of the formation of these metabolites when the incubation mixture
contained quinidine, which is a speci®c inhibitor of CYP2D6 (Kobayashi et al.
1989). Comparison of the amount of HO-NBA and HO-NEA formed (table 2)
indicated that the catalytic activity of CYP2D6 appeared not to be aåected by the
size of N-alkyl substituents; however, CYP2D6 activity was increased relative to
the extent of hydroxylation observed with AM. The formation of 4-HO-AM was
!
30 % of the quantities of 4-HO-NEA or 4-HO-NBA formed by the action of
CYP2D6 (table 2).
CYP2D6 catalysed the ring hydroxylation of N-alkylamphetamines, but its
contribution to this pathway was much less than for O-demethylation. As shown in
table 3, Clin for NBA hydroxylation was much lower than that of M-NBA O-
t
demethylation (2.81 versus 170.38 ll min for NBA and M-NBA respectively). In
}
addition, 3,4-dihydroxylated metabolites isolated in in vivo studies of NEA, NPA
and NBA metabolism were not detected in this study. This indicates that enzyme(s)
other than CYP2D6 is (are) involved in the formation of the catechol metabolites of
N-alkylamphetamines.
In summary, it has been established that M-NBA, M-NEA and M-AM are
excellent CYP2D6 substrates. It has also been shown that the CYP2D6-mediated
metabolism of all three substrates produces only one metabolite in each instance, the
corresponding O-demethylated products, HO-NBA, HO-NEA and HO-AM, and
the production of each is an eæcient process. It is also notable that the length of the
N-alkyl side chain in¯uences the activity of CYP2D6.
Unlike the O-demethylation of 4-methoxy-N-alkylamphetamines, CYP2D6
plays a minor role in the ring hydroxylation of NBA, NEA and AM. Enzymes other
than CYP2D6 are likely involved in the formation of ring hydroxylated metabolites
of N-alkylamphetamines.
No N-dealkylated products were isolated from any of the metabolism reactions
of N-alkylated-amphetamines or N-alkylated-4-methoxyamphetamines. Clearly,
CYP2D6 was incapable of catalysing the N-dealkylation of any of the examined
substrates.

CYP2D6 and in vitro metabolism
731
Acknow ledgem ents
The authors are grateful to the Alberta Mental Health Research Fund and the
Medical Research Council for funding.
References
lva n
rind
raffner
joqvist
, F., 1984, Relationship of N-demethylation of
A
, G., G
ami¯amine and its metabolite to debrisoquine hydroxylation polymorphism. Clinical Phar-
36
, M., G
, C. and S
macology and Therapeutics, , 515±519.
eckett idha
, K. K., 1974, The identi®cation of four metabolic products after incubation
B
C
C
, A. H. and M
of p-methoxyamphetamine with liver preparations of various species. Xenobiotica , , 297±311.
, J., 1976, The metabolism of amphetamines in mammals. Drug Metabolism Reviews,
219±280.
iramatsu
4
5
aldwell
,
ho
hu
istefano
hang
enden
, A. S. and J , D. J., 1990, Stereochemical
, A. K., H
, M., D
, E. W., C
diåerences in the metabolism of 3,4-methylenedioxymethamphetamine in vivo and in vitro : a
18
pharmacokinetic analysis. Drug Metabolism and Disposition, , 686±691.
umagai tefano ho
, A. K., 1996, Disposition of methylene-
i
C
C
,
T.,
K
,
Y., D S
,
E. W. and
C
dioxymethamphetamine and three metabolites in the brains of diåerent rat strains and their
possible roles in acute serotonin depletion. Biochemical Pharmacology, 51, 789±796.
oleman
llis
artin
ennard
ucker
, M. S. and T , G. T., 1996, 1-Methyl-4-
phenyl-1,2,3,6-tetrahydropyridine (MPTP) is N-demethylated by cytochromes P450 2D6, 1A2
, T., E
, S. W., M
, I. J., L
and 3A4 ± implications for susceptibility to Parkinson’s disease. Journal of Pharmacology and
277
Experimental Therapeutics,
outts ach
685±690.
aker
C
C
C
, R. T., B
expressed in a human cell line. Xenobiotica , , 33±47.
outts olaji aker
, M. V. and B
, G. B., 1997, Metabolism of amitriptyline with CYP2D6
27
, G. B., 1994, Metabolism of methoxyphenamine in vitro
22
u
, O. O., S , P. and B
, R. T., B
by a CYP2D6 microsomal preparation. Drug Metabolism and Disposition, , 756±760.
outts awson
, G. W., 1977, Urinary excretion of phenolic metabolites of N-(n-
,
R. T. and
propyl)amphetamine in man. Research Communications in Chemical Pathology and Pharmacology,
17
D
, 349±352.
awson
outts
ones
, G. R., 1978, In vivo metabolism of N-alkylamphetamines in
C
C
C
D
D
G
, R. T., D
the rat ± the eåect of N-alkyl chain length on oxidation of the aromatic ring. Research
22
, G. W. and J
Communications in Chemical Pathology and Pharmacology, , 589±592.
outts awson azakoff ong
, J. Y., 1976, In vivo phenolic metabolites
, R. T., D
, G. W., K
, C. W. and W
of N-alkylamphetamines in the rat. Evidence in favor of catechol formation. Drug Metabolism and
4
Disposition , , 256±261.
outts ones
, R. T. and J
2-(n-propyl-amino) propane, an intermediate in N-(n-propyl)amphetamine metabolism. Research
36
, G. R., 1982, In vivo and in vitro O-methylation of 1-(3,4-dihydroxyphenyl)-
Communications in Chemical Pathology and Pharmacology, , 173±176.
ybring lwin lm ndreasson yllenpalm
ahl
ertilsson
,
, M. L., T
, G., E
, C. E., A , C., A
L., 1994, Stereoselective disposition of mianserin is related to debrisoquin hydroxylation
56
, K., G
, M. and B
polymorphism. Clinical Pharmacology and Therapeutics, , 176±183.
uncan allstrom aulsen orman indeke ho
carbon substituents on the N-demethylation of N-methyl-2-phenethylamines by rat liver
a
-S , B. and C , A. K., 1983, Eåects of -
, J. D., H
, G., P
, U., L
11
microsomes. Drug Metabolism and Disposition, , 15±20.
eertsen oster ilson yr
asley
, W., 1995, Metabolism of
, S., F
, B. C., W
, D. L., C , T. D. and C
methoxyphetamine and 2-methoxyamphetamine in P4502D6-transfected cells and cell
25
preparations. Xenobiotica , , 895±906.
orrod
G
G
, J. W., 1973, The metabolism and excretion of ` amphetamines’ in man. In Frontiers in
Catecholamine Research edited by E. Usdin and S. H. Snyder (Oxford : Pergamon), pp. 945±950.
race
inter
acdonald
, T. L., 1994, Atypical metabolism of deprenyl and its
, J. M., K
, M. T. and M
a
enantiomer (S)-( 1)-N, -dimethyl-N-propynylphenethylamine, by cytochrome P450 2D6.
7
Chemical Research in Toxicology, , 286±290.
iratsuka hu istefano in
, A., C , T. Y., D , E. W., L , L. Y., S
chmitz
ho
, D. A. and C , A. K., 1995,
H
K
K
K
Inactivation of constitutive hepatic cytochromes P450 by phencyclidine in the rat. Drug
23
Metabolism and Disposition, , 201±206.
itchen remblay ndre ring
1979, Interindividual and interspecies variation in the metabolism of the hallucinogen 4-
dle
mith
illiams
, R. T.,
, I., T
, J., A
, J., D
, L. G., I
, J. R., S
, R. L. and W
9
methoxyamphetamine. Xenobiotica , , 397±404.
obayashi urray atson esardic
avies
oobis
, A. R., 1989, The
, S., M
speci®city of inhibition of debrisoquine 4-hydroxylase activity by quinidine and quinine in the rat
38
, S., W
, D., S
, D., D
, D. S. and B
is the inverse of that in man. Biochemical Pharmacology, , 2795±2799.
umagai in iratsuka arimatsu uzuki amada
guri
, K.,
, Y., L , L. Y., H
, A., N
, S., S
, T., Y
, H., O

732
CYP2D6 and in vitro metabolism
oshimura
ho
, H. and C , A. K., 1994, Participation of cytochrome P450±2B and -2D isozymes in
Y
the demethylenation of methylenedioxymethamphetamine enantiomers by rats. Molecular
45
Pharmacology, , 359±365.
umagai
(methylenedioxy)amphetamine and (methylenedioxy)methamphetamine by rat brain micro-
in
ho
,
L
,
L. Y.,
K
,
Y. and
C
A. K., 1992, Enzymatic and chemical demethylenation of
5
somes. Chemical Research in Toxicology, , 401±406.
asubuchi osokawa orie uzuki hmori
itada
arimatsu
, S.,
1994, Cytochrome P450 isozymes involved in propranolol metabolism in human liver microsomes.
M
, Y., H
, S., H
, T., S
, T., O
, S., K
, M. and N
The role of CYP2D6 as ring-hydroxylase and CYP1A2 as N-desisopropylase. Drug Metabolism
22
and Disposition, , 909±915.
idha
ooper
ailey
ubbard
,
M
,
K. K.,
methoxyamphetamine in dog, monkey and man. Xenobiotica , , 137±146.
arimatsu asubuchi osokawa hmori itada
C
,
J. K.,
B
,
K. and
H
J. W., 1981, The metabolism of 3-
11
uzuki
, T., 1994,
N
N
O
, S., M
Involvement of a cytochrome P4502D subfamily in human liver microsomal bunitrolol 4-
17
, Y., H
, S., O
, S., K
, M. and S
hydroxylation. Biological and Pharmaceutical Bulletin, , 803±807.
ielsen rosen ansen ram
, L. F., 1994, Single-dose kinetics of
,
K. K.,
clomipramine: relationship to the sparteine and S-mephenytoin oxidation polymorphisms.
55
B
,
K.,
H
,
M. G. and
G
Clinical and Pharmacological Therapeutics, , 518±527.
sutsui onzalez atoh asubuchi
no
orie
uzuki
, T. and
, S., T
, M., G
, S., 1995, Oxidative metabolism of bunitrolol by complementary DNA-expressed
human cytochrome P450 isozymes in a human hepatoma cell line (Hep G2) using recombinant
, F. J., S
, T., M
, Y., H
, T., S
arimatsu
N
5
vaccinia virus. Pharmacogenetics, 97±102.
hulgin argent aranjo
S
S
, A. T., S
psychotomimetics. Nature,
indrup rosen ansen
, T. and N
, C., 1969, Structure±activity relationships of one-ring
, 537±541.
aes orgensen
221
vero
ram
, L., 1993,
, S. H., B
, K., H
, M. G., A
-J
Pharmacokinetics of citalopram in relation to the sparteine and the mephenytoin oxidation
15
, T., O
, K. F. and G
polymorphisms. Therapeutic Drug Monitoring, , 11±17.
ring b
, L. G., 1970, Patterns of metabolism of -phenylisopropylamines in man and
mith
S
, R. L. and D
other species. In E. Costa and S. Garattini (eds), Amphetamines and Related Compounds;
Proceedings of the Mario Negri Institute of Pharmacological Research (New York : Raven), pp.
121±139.
u
ing
outts
aker
metabolites produced by isozyme CYP2D6 expressed in a human cell line. Xenobiotica ,
aneshtalab
, G. B. and D , M., 1993, Analysis of imipramine and three
S
, P
, C
, R. T., B
23
,
1289±1298.
seng
enon
oh
,
T
T
,
L. F.,
amphetamines on the release and uptake of biogenic amines in brain tissue. Journal of
197
M
,
M. K. and
L
H. H., 1976, Comparative actions of monomethoxy-
Pharmacology and Experimental Therapeutics,
, 263±271.
ucker
ennard
D. A. and
llis
oods
ho
in
iratsuka
, A.,
, G. T., L
chmitz
, M. S., E
, S. W., W
, H. F., C , A. K., L , L. Y., H
hu
, T. Y., 1994, The demethylenation of methylenedioxymeth-
S
,
C
47
amphetamine (` ecstasy ’) by debrisoquine hydroxylase (CYP2D6). Biochemical Pharmacology,
1151±1156.
,
illiams
aldwell
ring
D ,
amphetamines in various species. In Frontiers in Catecholamine Research, edited by E. Usdin and
W
,
R. T.,
C
,
J. and
L. G., 1973, Comparative metabolism of some
S. H. Snyder (Oxford : Pergamon), pp. 927±932.
asui
tani
ugawara
aneko
hkubo
sanai
shida
ihara
ondo
, T.,
Y
, N., O
S
,K., K
, K. and F
, S., O
, T., O
, Y., 1995, Inhibition of trazodone metabolism by thioridazine in
17
, T., I
, M., M
, K., K
ukushima
humans. Therapeutic Drug Monitoring, , 333±335.